Campaign for Cambridge Freedoms

Analysis of the Vice-Chancellor's Proposal

Freedoms at Cambridge University are under threat. A policy change proposed by the University's Vice-Chancellor will severely undermine academic freedom, adversely change the employment conditions of existing faculty by expropriating our copyrights and patents, make the recruitment of new faculty significantly harder, and kill the goose that lays the golden eggs - the large number of high-tech businesses spun off from the University, known as the Cambridge Phenomenon.

The proposal was discussed in the Regent House on the 15th and 22nd October.

This page presents an analysis of the proposed changes. It was written not just for people preparing to speak at the Regent House, but for people anlysing the effect that the proposals will have on your academic life (or your business). I explain what they will do, give pointers to the relevant economic literature, and estimate the main positive and negative effects of the changes for individual faculty members, research students, departments, the University, the regional economy, and academic freedom.

The analysis is broken down under the following headings:

See also the main web page of the Campaign for Cambridge Freedoms; other sources include Mike Clark's page on the history of intellectual property at Cambridge, the original page I wrote on this subject, and press coverage in the Observer, the Telegraph, the Cambridge Evening News, ZDNET, THES, the Wall Street Journal and Frankfurter Rundschau. There's also the BBC, which our Vice Chancellor told that "The university has a right to a share because I think there are very few true individuals. Most people have to rely on others". (If he is asserting that the median Cambridge faculty member has never published a significant single-author paper, I'd like to see his statistics.)

There's also a rather appropriate Dilbert, which you can buy on a T-shirt.

(: I may update this page once the Cornish committee reports. Meanwhile, some new research has been published in a forthcoming special issue of the International Journal of Industrial Organization on the Economics of Intellectual Property at Universities and some more in the January 2003 issue of the Journal of Technology Transfer - note that this issue is free online. There are also nice short and long literature surveys from Ajay Agrawal.)

Copyright

Previously, Cambridge academics owned the copyright in everything we produced - research papers, lecture notes, books, software, databases, drawings, music and the performance of other people's creations. Under the new regime, the University will assert ownership of all IPR, including all copyright, with a rather narrow exception: `The University will not claim ownership of copyright in normal academic forms of publication including books, articles, and lectures, or other similar works generated by staff, unless those works have been commissioned by a sponsor or by the University'.

It is most unusual for universities to claim copyright in the material created by their faculty or students, with the exception of some very restricted types of material like exam papers. In the UK, there was an attempt in 2000 by UCL to do this, which was defeated by widespread resistance from faculty members. Some US universities have tried to claim rights to lecture notes, but this has also been strongly resisted and I am aware of no respectable universities that do it. (See for example the practices at California and Cornell, and the policy statement of the American Association of University Professors.) Cambridge would become a laughing stock if we let our bureaucrats grab rights that are the fiercely defended preserve of their creators everywhere else.

The justification given privately by senior members of Cambridge University for this attempted grab is that an engineering lecturer used tens of thousands of pounds' worth of resources from our Audio Visual Aids Unit to create a fancy course, which he then sold privately as well as giving to local students. It seems that the only way the Vice Chancellor can see to prevent such abuses is to own our copyright. (Powerpoint made the AVA obsolete, and it was disbanded a year or two ago, so the point is now irrelevant anyway.)

The consequences of giving the University default ownership of copyright will be grave.

As for the economic effects, we know quite a lot about software. A few years ago, the University of California tightened the rules on faculty making software available for free; the Regents have been persuaded to relax them again, as the net effect was a reduction in research funding from industry. The reasons for this are complex, but are starting to be understood. Many industrial sponsors want work they've paid for to become a standard, while a lot of the brightest programmers are reluctant to work on closed systems, at least for the sort of wages Universities pay to research staff. There is a strong signalling effect: working on systems whose code will be published enables a programmer to use this code as a reference when seeking fully paid jobs later. There is also an element of idealism, and the satisfaction of contributing to a product like GNU/linux or Apache that is very widely used. The economics of open source software is now a specialist subject in its own right; for more, see the relevant conferences.

There is little economic literature speaking directly to the likely effects of the university's copyright grab as a whole, as it is so broad as to be unprecedented. The normal copyright rule, that one finds at places like the University of California, is that unpatentable scholarly works are the property of the originator. But there are two sets of general results that have some relevance.

It is extremely doubtful that the copyright grab will generate a significant amount of money. Our technology transfer office is geared toward royalties from patents rather than copyrights, and the small number of high-value copyright earners can easily avoid sharing their royalties (in Hawking's case, he is already 60, so he could take early retirement). Celebrities whose personal papers are worth serious money will presumably make them over to family trusts before the deadline. So I see no upside for the university's finances. On the other hand, there is a huge downside, not just for faculty in disciplines like computer science, physics and music where the new regulations will be oppressive, but as a general dampener on faculty esprit de corps and on the university's ability to recruit and retain world-class staff. As well as the economic issues, there are control issues; there will be a further chilling effect from the prospect that the university might sell my course material for use in an institution of which I disapprove. Finally, if the proposed reforms do in fact pass a vote in the Regent House, they will make the University of Cambridge a laughing stock.

Patent rights

The first wave of debate over the proposal has focused on patent rights. RSD claims that the university could increase its patent income if it were more aggressive, while opponents of the measure have pointed out that replacing the current laissez-faire arrangements with a rigid bureaucracy would have serious effects for Cambridge's success at technology transfer.

The debate has been confused by the fact that there are three separate types of patent arrangement at present in the university (as indeed at MIT, Stanford, UMIST and many other institutions).

The main effect of the proposed policy change on patent rights is that in future independent research will be treated in the same way as externally funded research, that is, the university will own the rights by default. It is this that caused much of the initial furor. Although the government/industry/independent split is about 60-20-20 across higher education as a whole, Cambridge has a much higher proportion of private sponsorship and of independent research, and in some disciplines (such as computer science), most patentable ideas arise independently. In others, external funding is non-commercial or even anti-commercial: our Sanger Centre got huge charitable funding to publish the human genome before the Americans could patent it.

What will be the likely economic effects?

There is a substantial economic literature in the USA studying the effect of the Bayh-Dole Act of 1980 which gave universities the right to patent and licence the results of government-funded research. The most important papers appear to be the following.

This literature demolishes the argument of Research Services Division that the link between the liberal IP rules at MIT, Stanford and Cambridge and our success at technology transfer is `superficially attractive, but is unprovable and a policy based on assertion and belief is hard to justify'.

There is less literature specific to Cambridge, but this is compensated by quality: the book by Segal Quince & Wicksteed that coined the phrase `Cambridge Phenomenon' for the large number of high-tech companies clustered around the university. It argues that an ethos of self-confidence of the University of Cambridge inspired start-up companies around it, and that the ability to retain intellectual property rights allowed university people to try exploiting their ideas with new firms.

As an example of how IP at Cambridge works, I got the idea of using balanced self-checking logic for smartcard CPUs in a tearoom conversation with my colleague Simon Moore in early 1999. We filed a patent application using our own money and proceeded to raise seven-figure financing for a collaborative research project that, from the scientific point of view, was outstandingly successful. Our paper on this technology won the best presentation award at Async 2002, and en route we discovered a new semiconductor testing technology that made the front page of the New York Times. From the business point of view, the timing was unlucky; if we'd started two years earlier we'd be rich. But that's life: most technologies fail, so technologists have to take risks, and individual faculty are more likely to take them than institutions - so long as the incentives are there. Without at least the prospect of creating something valuable, it's unlikely that we'd have put in all the effort that we did, and made the contribution to the basic science that we ended up making.

Our experience, and that of colleagues, is that our `day' job of research and our `weekend' activities of technology transfer, consultancy and so on, are not substitutes, but complements. Ideas from research feed into business, and vice versa. There can be no telling, at the start of a project, whether it will come up with an idea that will win a Nobel Prize, or found a large company. In the case of our smartcard idea, an attempt to make money led to scientific progress instead; it often goes the other way too. Science is an unpredictable business, and erecting barriers in the way of exploitation will inevitably hinder research too.

In addition to the IP polices at MIT and Stanford, and our current policy, you might care to have a look at the policies of other leading US universities, such as CMU and Columbia. CMU gives faculty all their copyright and 85% of patent income, unless the research was funded by the university in which case they get 50%; Columbia gives faculty all their copyright and 25% of patent, plus a further 25% to be spent on their research, and allows faculty to veto any license on conscientious grounds. The worst is Johns Hopkins where the inventor's share is as low as 10%. (Johns Hopkins is notoriously the worst major US university when it comes to technology transfer.) Some other US policies are linked here, while the policies of Canadian universities are here.

The view of business must surely matter too. After all, they are the customers to whom RSD hopes to sell the IP that will magically appear once we're compelled to hand it over. The view of local high-tech business people is uniformly hostile to the proposals, and they have recently written a letter to the Vice Chancellor expressing this (the text of which I'm unfortunately not allowed to post here).

Other intellectual property rights and restrictions

Patent and copyright do not exhaust intellectual property. The proposed policy says that `IPR arising from intellectual property devised, created, or made on after 1 January 2003 will be owned by the University' and defines IPR as `all rights arising from intellectual property devised, created, or made by staff in the course of their employment by the university whether those rights are capable of registration or not, the most common forms arising from academic research being patents, know-how, confidential information, and copyright in the form of designs, software, and associated documentation.' The effect of this will be broad and complex. Problems that we have already identified include the following. These assorted intellectual property rights will have a range of effects. The database right will make it harder for many academics to change jobs, which will exert downward pressure on salaries, as the university will not have to pay as much attention to the need to retain staff. The loss of control over the uses to which our ideas are put, and the diminution of effective redress in some research ethics disputes, will decrease the attraction of Cambridge as an employer and make it harder to recruit good faculty.

There is also the simple point that academia is in many ways a gift economy; people acquire status in proportion to the benefits that we give to our peers. At the formal level, such gifts take the form of new facts and theories on which others can build. At the informal level, they are backed up by numerous other kinds of gift - helping organise a conference, running a website with a specialist subject bibliography, maintaining a critical software tool, or providing informal free consulting to colleagues in other disciplines. All these activities are under threat, directly or indirectly, from the proposed policy. How can you participate in a gift economy when the things that you can give are under the control of others?

Consulting arrangements by faculty members

The director of research services has said in an email that `in the near future the Research Policy Committee will be considering the whole question of consultancy in the University'.

At York, you have to hand over 40% of your consulting income to the university; at Glasgow, it's 50%. I wonder what rate will be set here?

If RSD is determined to lock down our patents and copyrights, then it's logical for them to lock down consultancy too.

Making appeals to greed and spite is easy; `your income is taxed so theirs jolly well ought to be too.' Resisting the temptation is important. What will happen to the university if economists lose their consultancies with banks in the City, if lawyers have to hand over much of their earnings from pleading cases at the Bar, and medics find their private practice closed down? The result would be a Cambridge consisting of humanities scholars, a few committed left-wingers, and a lot of third-raters. Is that where you really want to go? To lose all the profitable faculties like computer science and law, and be left with the loss-making bits?

Organisational issues

There are a number of serious organisational issues that need to be tackled. The merger of our technology transfer organisation with the administration of research grants creates some acute conflicts of interest; for example, a tech transfer associate can pressure a faculty member into handing over more equity in a project with the threat of interference in funding streams. There is also the conflict between the mission to provide technology transfer advice, and the goal of generating revenue for the university.

For this reason, I argued in the Times Higher Educational Supplement that we need internal competition in the university for technology transfer. Where a patent must by law be held by the university, the inventor should have a choice of teams to which she can turn for offers. Colleges provide one natural focus for teams: some of them, like Trinity and Johns, have world-famous science parks. Some departments are active too. The technology transfer scene at Cambridge is already too large and diverse, with competing companies and clashing egos, to be squeezed into a single bureaucratic mold. We also need to think about whether we need to have a separate place to which inventors can turn for independent advice.

It's also important to understand that technology transfer is not a panacea, which will overnight fix the university's deficit. On the contrary; it's very easy to lose money. Of the University of California campuses, three lost money at this business in 2001 (Santa Barbara lost $415,000, Santa Cruz lost $212,000 while San Diego lost a whopping $1,904,000).

Recall that of the studies cited above, Henderson, Jaffe and Trajtenberg reported that the rush to patent did not unlock any treasure trove of new valuable IP, but just added a lot of junk patents to the few good ones that came along anyway; Jaffe and Lerner emphasised the harm done by bureaucratisation of the process; while Thursby and Thursby showed that TTOs have a steep learning curve to climb. RSD, being new to the game, might take a decade to learn it; the returns to investment in this business take typically ten years to arrive (most of the university's patent royalties today are from the Campath patent, filed in 1987); and in the intervening twenty years, RSD's TTO staff will consume perhaps one or two million pounds a year.

RSD say that they need control of IP in order to `do deals', and their supporters have quoted a deal at Oxford whereby the chemistry department sold their next 15 years' inventions to a venture capital group in return for some money towards a new building. There is a small problem here. The Oxford deal was done with the consent of faculty members; but when RSD has tried to do such deals, the affected faculty have not even been told! One disgraceful example was when Microsoft was planning to open a research lab in Cambridge a few years ago, and RSD suggested to them that they could get first refusal on all IP generated by computing faculty. We, the faculty, were never told; and had we been told, we'd probably have run Microsoft out of town on a rail. As it was, the deal was held up for eighteen months while wiser heads worked to persuade the top people at Microsoft that RSD's promise had been empty. When a delicately negotiated compromise was presented to us at a staff meeting, the then head of department was roasted, and Microsoft had to do without any agreement at all. After that, how can faculty members trust RSD? Would you buy even a used car from someone who's previously tried to sell the Brooklyn Bridge?

There is a widespread suspicion that RSD wants to do an even larger deal than that. The precedent that's being quoted is that of Imperial which has sold an option on all its future IP for an eight-figure sum. However much the Vice-Chancellor might like a cheque for thirty or even fifty million pounds, to cover for a few years the trading deficit into which his administration has led us, it seems most imprudent to conduct such an asset sale under distress conditions at the nadir of the technology market. Prospective purchasers will surely become aware that the sale is contrary to the wishes - and the rights - of the people who're supposed to generate the IP they're paying for. The most likely outcome is a paltry sum of money followed by a lot of litigation.

There is also the more basic point that RSD is incompetent. We all suffer from their mistakes and delays with the basic paperwork of research grant administration; last year, their inability to prepare a contract in time cost us an eight-figure research investment from Marconi. Their new wheeze, of building a vast new empire based on our IP, will merely distract them from the urgent task of getting the basics right.

Basic Rights

Undermining intellectual property will have more subtle effects. IP rhetoric is often used to protect intangible goods and hygiene factors such as cognitive property, status, independence, the right to self-governance, and the boundaries of particular academic communities. This has been (justly) criticised by thinkers such as Stallman and McSherry: using IP rights language risks losing many of the nuances of the language of academic rights. Indeed, there are some practical - and serious - consequences that the proposed regulations will have for academic freedom.

Action

We must defeat this measure. We must assert our continued ownership of all intellectual property except patents tied to research contracts, and these patents must be prised free from the dead hand of RSD. That means introducing internal competition into technology transfer.

There was a Discussion in the Regent House on the 15th and 22nd October, that was even reported by the BBC. One of the highlights was a speech by Tom Körner; we also received overwhelming support from Cambridge University Students' Union.

According to our spies, the Council referred the matter back to the research Policy Committee, but as the secretary of this committee is the Director of Research Services (the main advocate for the policy within the bureaucracy) we can expect their next recommendation to be similarly unacceptable. Most likely, they will try to get their way piecemeal - take the patents this year, the database rights next year, the copyrights the year after that, and then finally impose a tax on all our consultancy on the grounds of being fair to those who've already lost most of their royalty income. In this case, we'll have to hang together, or they'll hang us separately.

At the Council elections in November, both of the members elected in class b were pro-freedom (Martin Rees and me) while two of the four candidates elected in class c had also come out in favour of freedom (Richard Barnes and James Matheson) and the other two had declined to take a position. The openly pro-RSD candidates - John Bell and Jeremy Sanders - were both defeated. The mood of the University seems clearer on this issue than on just about any other.

Ross Anderson

The report of the Cornish Committee was published on the 6th August 2003. This committee was set up by the Vice-chancellor to advise him how to proceed, and its report suggested much as I predicted two paragraphs above: take patent rights now, as it wouldn't be politically possible to do any more for the time being. Nonetheless, it is still a fairly damning indictment of how the University has managed IP over the last couple of years.

Acknowledgements: a number of people have contributed to this analysis and exposition including Michael Ashburner, James Boyle, Mike Clark, Simon Deakin, Richard Evans, Chris Lowe, David MacKay, David Newbery. Pam Samuelson and a number of colleagues who are still in what VCs call `stealth mode'.